Bromodomain-containing Protein 4 Activates Voltage-gated Sodium Channel 1.7 Transcription in Dorsal Root Ganglia Neurons to Mediate Thermal Hyperalgesia in Rats

EPIGENETIC mechanisms at dorsal root ganglia (DRG) neurons are relevant for the development of chronic hyperalgesia.¹  For example, histone acetylation influences the expression of nociceptive genes in DRG neurons during chronic hyperalgesia states.^2,3  Bromodomain-containing protein 4 (Brd4), which is a member of the bromodomain and extraterminal (BET) protein family, is recognized as a key epigenetic reader because Brd4 binds acetylated promoter histones and promotes transcriptional coactivation and elongation.^4,5  In hippocampal neurons, Brd4-associated modifications of gene transcription are associated with learning and memory⁶ ; in contrast, pharmacologic interference with Brd4 functions results in memory deficits in mice.⁶  Notably, in addition to affecting learning- and memory-associated plasticity in certain brain areas, epigenetic mechanisms regulate nociception-related plasticity in the DRG.⁷  Moreover, JQ1, which is a BET inhibitor, inhibits inflammation by attenuating the recruitment of Brd4 to the promoters of an inflammatory gene⁸  that, in the DRG, participates in inflammatory hyperalgesia.⁹  Nevertheless, the potential contribution of Brd4-associated epigenetic mechanisms in the DRG to the progression of inflammatory hyperalgesia has not been established.

Brd4 is described as a positive regulatory component of cyclin-dependent kinase (CDK)-9 (CDK9) for RNA polymerase II (RNAPII)–dependent transcription.^10,11  Brd4 is necessary for the formation of transcriptionally active CDK9, the recruitment of CDK9 to acetylated promoters, and the phosphorylation of RNAPII to activate the transcription of inflammatory genes.¹²  Interestingly, histone acetylation at regulatory sequences in the voltage-dependent sodium channel (Nav) gene is associated with chronic hyperalgesia-associated Nav expression in DRG neurons.¹³  The upregulation of Nav1.7 in DRG neurons has been implicated in the induction and maintenance of chronic inflammatory hyperalgesia.¹⁴  The inhibition of Nav1.7 using a monoclonal antibody effectively suppressed inflammatory hyperalgesia in mice.¹⁵  These observations prompted us to investigate whether Brd4 could epigenetically modify the expression of Nav1.7 in DRG neurons to mediate the development of inflammatory hyperalgesia by activating CDK9-dependent RNAPII phosphorylation.

Tumor necrosis factor-α (TNF-α) participates in the development of inflammatory hyperalgesia in the DRG.¹⁶  Brd4 is an important regulator of TNF-α–induced inflammatory gene transcription/expression.¹⁷  Notably, TNF-α stimulates the association of Brd4, CDK9, and phosphorylated (p)RNAPII with target genes in inflammatory environments¹⁸ ; in contrast, JQ1 suppresses the expression of TNF-α–associated inflammatory genes.¹⁹  In particular, TNF-α induces the expression of Nav1.7 in DRG neurons to regulate the development of chronic hyperalgesia.²⁰  Collectively, this evidence prompted us to hypothesize that TNF-α may impact Brd4-regulated Nav1.7 transcription in DRG neurons, thus supporting inflammatory hyperalgesia.

Adult male Sprague–Dawley rats weighing 200 to 250 g were used in this study. All of the animals were housed at room temperature (23° + 1°C) with a 12-h light–dark cycle (lights on 8:00 am to 8:00 pm) and fed food and water ad libitum. The surgical procedure and experimental protocols performed in this study were conducted in accordance with the guidelines of the International Association for the Study of Pain²¹  and were reviewed and approved by the institutional review board of Taipei Medical University (Taipei, Taiwan). The animals were randomly allocated to groups using a Research Randomizer (available at https://www.randomizer.org/; accessed February 5, 2015), and the sample size of each group was based on our previous experience. In each group, seven rats were used for the behavioral tests and immunohistochemistry; six rats were used for the quantitative reverse-transcription polymerase chain reaction (PCR), Western blotting, and chromatin immunoprecipitation–quantitative PCR (ChIP); four to six rats were used for the electrophysiologic analysis; and eight rats were used to assess the rearing behavior. The investigators were blinded to the treatment groups in all of the experiments. Eleven rats showed neurologic deficits after the catheter implantation and were excluded from the statistical analysis. Sex differences in the outcomes measured in this study are possible but were not explored in the present work.

The complete Freund’s adjuvant (CFA) rat model was established as described previously.^22,23  In brief, the rats were deeply anesthetized with isoflurane (induction 5%, maintenance 2% in air; Baxter, Puerto Rico) and received a subcutaneous injection of 100 μl CFA (1 mg/ml; Sigma-Aldrich, China) or saline into the plantar side of the left hind paw. At 3 h and 1, 3, 5, and 10 days after the CFA or saline administration, a thermally evoked paw-withdrawal response was assessed using a Hargreaves-type device (model 336 combination units; IITC/Life Science Instruments, USA). Briefly, the rats were acclimated to the test chamber for 30 min on a glass surface maintained at 25°C. Radiant heat generated by a halogen projection bulb was used to stimulate the hind paw. An abrupt withdrawal of the paw was detected using a motion sensor, which triggered the termination of the stimulus, and the paw withdrawal latency (PWL) was automatically determined. A cutoff of 20 s was used to avoid tissue injury. Motor function was assessed using an accelerating Rotarod apparatus (LE8500, Ugo Basile, Italy). For acclimatization, the animals were subjected to three training trials at 3- to 4-h intervals on 2 separate days.²⁴  During the training sessions, the rod was set to accelerate from 3 to 30 revolutions per minute over a 180-s period. During the test session, the performance times of the rats were recorded until the cutoff time of 180 s. Three measurements were obtained in 5-min intervals and averaged for each test.

Each rat was placed in the center of a black arena (50 × 50 cm, 40 cm high). A video camera system (KMS-63F4; AWON, Taiwan) was used to record the activity of the rats for 30 min. Locomotor activities, including the traveling distance (in centimeters), rear durations, and number of rears were analyzed using the EthoVision XT tracking software (Noldus Information Technology, The Netherlands).

The dissected DRG (L5 to L6) samples were homogenized in 25 mM Tris-HCl, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, and 0.1% sodium dodecyl sulphate (SDS) supplemented with a complete protease inhibitor mixture (Roche, Germany). After incubation on ice (1 h), the lysates were centrifuged (14,000 revolutions per minute, 20 min, 4°C). All of the protein concentrations were determined using a bicinchoninic acid assay. The supernatant (50 μg) was separated on an acrylamide gel and transferred to a polyvinylidene difluoride membrane, which was then incubated (1 h, room temperature) with rabbit anti-Brd4 (1:1,000, Abcam, USA), rabbit anti-Nav1.7 (1:200, Alomone Labs, Israel), rabbit anti-CDK9 (1:1,000, Abcam), rabbit anti-pRNAPII (1:1,000, Abcam), or mouse anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 1:4,000, Santa Cruz Biotechnology, USA) antibodies. The blots were washed and incubated (1 h, room temperature) with peroxidase-conjugated goat antirabbit immunoglobulin G (IgG; 1:8,000, Jackson ImmunoResearch, USA) or goat antimouse IgG (1:8,000, Jackson ImmunoResearch) antibodies. The protein bands were visualized using an enhanced chemiluminescence detection kit (ECL Plus, Millipore, USA) and then subjected to a densitometric analysis using Science Lab 2003 (Fuji, Japan).

Coprecipitation was performed as described previously.^25,26  The dorsal horn extracts were incubated with a rabbit polyclonal antibody against Brd4 (rabbit, 1:1,000, Abcam) overnight at 4°C. A 1:1 slurry protein agarose suspension (Millipore) was added to this protein immunocomplex, and the mixture was incubated (2 to 3 h, 4°C). The agarose beads were washed once with 1% (volume/volume) Triton X-100 in the immunoprecipitation buffer (50 mM Tris-HCl [pH 7.4], 5 mM EDTA, and 0.02% [weight/volume] sodium azide), twice with 1% (volume/volume) Triton X-100 in the immunoprecipitation buffer (300 mM NaCl), and thrice with only the immunoprecipitation buffer. The bound proteins were eluted in SDS–polyacrylamide gel electrophoresis sample buffer at 95°C. The proteins were separated on SDS–polyacrylamide gel electrophoresis, electrophoretically transferred to polyvinylidene difluoride membranes, and detected using primary antibodies against Brd4 (rabbit, 1:1,000, Abcam), CDK9 (rabbit, 1:1,000, Abcam), and Nav1.7 (rabbit, 1:200, Alomone Labs).

After perfusion (100 ml phosphate-buffered saline [PBS], followed by 300 ml 4% paraformaldehyde in PBS; pH 7.4), the DRG samples were harvested (L5 to 6), postfixed (4°C for 4 h), and cryoprotected overnight in a sucrose solution (30%). The DRG sections (16 μM) were cut using a cryostat and processed for immunofluorescence. To assess the expression of Brd4 in the DRG, the DRG sections were incubated overnight (4°C) with a rabbit anti-Brd4 antibody (1:400, Abcam). To investigate the distribution of Brd4 in the DRG, the DRG sections were incubated overnight (4°C) with a mixture of rabbit anti-Brd4 (1:400, Abcam) and mouse anti-NeuN (a neuronal marker; 1:500; Millipore), mouse anti-glial fibrillary acidic protein (a marker of satellite cells; 1:1000; Millipore), mouse anti-ionized calcium binding adaptor molecule 1 (anti-Iba-1, a marker of macrophages; 1:400; Santa Cruz Biotechnology), neurofilament 200 (NF-200, a marker of A-fiber afferents and medium–large myelinated neurons; 1:1,000; Sigma-Aldrich), isolectin B4 (IB4, a marker of small nonpeptidergic C-nociceptors; 1:400; Sigma-Aldrich), or calcitonin gene-related peptide (CGRP; a marker of small peptidergic C-nociceptors; 1:1,000; Sigma-Aldrich) antibodies. After three rinses with PBS, the sections were then incubated (1 h, 37°C) with Alexa Fluor 488 (1:1,500; Invitrogen, USA) and Alexa Fluor 594 (1:1,500; Invitrogen). To examine the interactions among Brd4, CDK9, pRNAPII, and Nav1.7, the specific antibodies were mixed with 10X reaction buffer (Mix-n-Stain; Biotium, USA) at a ratio of 1:10. The solution was then transferred to a vial containing dye (CF, Biotium) and incubated in the dark (30 min, room temperature). The DRG sections were sequentially incubated (overnight, 4°C) with the diluted solutions, that is, rabbit anti-Brd4 (1:400, Abcam), rabbit anti-Nav1.7 (1:500, Alomone Labs), rabbit anti-CDK9 (1:500, Abcam), and rabbit anti-pRNAPII (1:500, Abcam) antibodies, and washed five times between each incubation. The DRG sections were subsequently rinsed in PBS, and coverslips were applied. After excitation, the fluorescent markers were easily detected using a camera-coupled device (X-plorer; Diagnostic Instruments, Inc., USA) using fluorescence microscopy (LEICA DM2500, Leica Camera, Germany). For the quantifications, the 20th section (L5 to L6 DRG samples) was selected from a series of consecutive DRG sections, four sections were counted for each DRG, and seven rats were analyzed in each group. To determine the percentage of labeled neurons, the number of positive neurons was divided by the total number of neurons.

Briefly, the dissected DRG (L5 to L6) was quickly removed, completely submerged in a sufficient volume of RNAlater solution (AM7021; Ambion, USA) overnight at 4°C to enable thorough penetration of the tissue, and then maintained at 80°C. Total RNA was isolated under RNase-free conditions using RNA isolation kits (74106; Qiagen, USA). Reverse transcription was performed using complementary DNA reverse transcription kits (205311; Qiagen). Real-time PCR was performed on a 7500 Real-Time PCR system (Applied Biosystems, USA). TaqMan Universal PCR Master Mix (2X) and gene expression assay probes for the target genes GAPDH (Rn99999916_s1, Applied Biosystems) and Nav1.7 (Rn.PT.58.34869054, IDT, USA) were used. The reactions (total volume, 20 µL) were incubated at 95°C for 20 s, followed by 40 cycles of 1 s at 95°C and 20 s at 60°C. The relative messenger RNA (mRNA) levels were calculated using the 2(-delta delta cycle of threshold [Ct]) method.²⁷  All of the Ct values were normalized to GAPDH. Nontemplate controls, minus-reverse transcriptase controls, and standard curves were used to evaluate the specificity and efficiency of the Nav1.7 (scn9a) quantitative reverse transcription PCR quantification.

ChIP was performed using a ChIP kit (Millipore) according to a modified protocol by the manufacturer. The dissected DRG samples were cut into small pieces (1 to 2 mm³) using razor blades. The minced samples were treated with fresh 1% paraformaldehyde in a PBS buffer by gentle agitation for 10 min at room temperature to cross-link the proteins to the DNA. Then, the tissues were washed and resuspended in lysis buffer, the lysates were sheared by sonication to generate chromatin fragments with an average length of 200 to 1,000 base pairs, and 1% of the sonicated chromatin was saved as an input control for the quantitative PCR. The chromatin was then immunoprecipitated for 2 h at room temperature with rabbit anti-Brd4 (5 μg, Abcam), rabbit anti-CDK9 (5 μg, Abcam), rabbit anti-pRNAPII (5 μg, Abcam), and rabbit anti-H3 (5 μg, Millipore) antibodies or an equivalent amount of rabbit IgG (5 μg, Sigma-Aldrich). The protein–DNA immunocomplexes were precipitated overnight using protein G magnetic beads at 4°C. After the beads were washed, they were resuspended in the ChIP elution buffer, incubated with proteinase K at 62°C for 2 h, and then incubated at 95°C for 10 min to reverse the protein–DNA cross-links. The ChIP signals were quantified via a quantitative PCR analysis on a 7500 real-time PCR system (Applied Biosystems). The following specific primer pairs were used to amplify the Nav1.7 (scn9a) promoter region: 5′-CTGCGGAAGGAAGAAATCAG-3′ and 5′-TCTCGTGCTTCAAACTGTGG-3′.

The 19-nucleotide small-interfering RNA (siRNA) duplex molecules used to target Brd4 and CDK9 were 5′-GAUGAAGCCUGUAGAUGUA-3′ and 5′-GCGAUGAGGUCACCAAGUA-3′, respectively, and the missense nucleotide sequence was 5′-UGAUAUUACCCUGAAUAUG-3′. The missense or siRNA construct was intrathecally administered using a polyethyleneimine (10 μl, Dharmacon, USA)-based gene-delivery system. The Brd4 mRNA-targeting siRNA or missense siRNA was intrathecally injected daily from day 4 to day 1 before the saline or CFA injection.

JQ1 (an inhibitor of BET binding to acetylated histones; 10 μl; 10, 30, and 100 μM; Cayman Chemical, USA; dissolved in 1% dimethyl sulfoxide, intrathecal) and the TNF-α–neutralizing antibody (10 μl; 10, 30, and 100 ng, R&D Systems, Inc., USA, intrathecal) were administered to the rats via multibolus intrathecal injections. At 3 h after the intraplantar saline/CFA injections, the animals were intrathecally injected with JQ1, a vehicle solution, or the TNF-α–neutralizing antibody once every 6 h for 18 h. The animals were examined 3 h after the final JQ1–vehicle solution–antibody injection. TNF-α (10 μl; 1 pM; Sigma-Aldrich) was administered intrathecally by a bolus injection, and subsequent experiments were performed 3 h after the injection. A vehicle solution was administered at a volume identical to that of the tested agents to serve as a control. For the intrathecal injections, a lumbar puncture was made at the L5 to L6 level using a 30-gauge needle under brief isoflurane anesthesia.²⁸ 

Male Sprague–Dawley rats were deeply anesthetized with 2 to 3% isoflurane. A laminectomy was performed, and the ipsilateral side of the L5 and L6 DRG was removed. The dissociated DRG tissues were immediately transferred to Hank’s balanced salt solution (pH 7.3 to 7.4). The attached nerve roots of the DRG were trimmed, and the epineurium and perineurium that form the capsule were carefully removed as described previously.²⁹  The DRG tissues were incubated in 2 ml Hank’s balanced salt solution containing type I collagenase (1.5 mg/ml, Sigma-Aldrich) at 37°C for 40 min. After enzymatic digestion, the DRG tissues were transferred to a holding chamber containing artificial cerebral spinal fluid (117.0 mM NaCl, 4.5 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 1.2 mM NaH₂PO₄, 25.0 mM NaHCO₃, and 11.4 mM dextrose) and oxygenated with 95% O₂–5% CO₂ (pH 7.4).The currents of the intact DRG neurons were recorded as described previously.³⁰  The extracellular solution was composed of 140 mM NaCl, 3 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, 10 mM HEPES, 5 mM CsCl, and 20 mM tetraethylammonium chloride (pH 7.3, adjusted with caesium hydroxide). The glass recording pipettes (3 to 5 MΩ) were filled with an internal solution composed of 110 mM CsF, 25 mM NaCl, 1 mM CaCl₂, 10 mM EGTA, 20 mM tetraethylammonium, 10 mM glucose, and 10 mM HEPES (pH 7.3; adjusted with caesium hydroxide). Electrophysiologic signals were acquired using an Axon setup (Molecular Devices/Axon Instruments, USA). The signals were sampled at 5 to 10 kHz using pCLAMP 9.2 (Molecular Devices, USA) with an amplifier (Axopatch 200B, Molecular Devices) and an AD-converter (Digidata 1322A, Molecular Devices) and analyzed using Clampfit 9.2 (Molecular Devices). The total whole-cell Na⁺ current was recorded using a series of depolarizing voltages from –60 to 0 mV (40-ms pulse duration) in 5-mV increments at 5-s intervals. To record the Nav1.7 currents, the currents were isolated using a selective Nav1.7 channel blocker (ProTx-II; Sigma-Aldrich). The ProTx-II–sensitive Nav1.7 currents were derived by subtracting the Nav currents that were recorded in the presence of 5 nM ProTx-II from the total Nav currents. The L5 and L6 ganglia were collected from the rat, and two slices were dissected from each ganglia. Only 1 neuron in each slice was recorded for the electrophysiologic analysis, which was performed 2 to 6 h after the tissue dissection.

All of the data in this study were analyzed using Sigma Plot 10.0 (Systat Software, USA) or Prism 6.0 (GraphPad, USA), and the results are expressed as the mean ± SD. A two-way ANOVA was used to assess changes in the values of serial measurements over time, and a Tukey post hoc test was used to compare the means of the groups (fig. 1, A and B). Paired two-tailed Student t tests were used to compare the means between the groups (fig. 1C). Statistical comparisons were performed using one-way ANOVA, followed by Bonferroni corrections for the post hoc analysis (fig. 1, D through G). Statistical comparisons were performed using one-way ANOVA, followed by Tukey tests for the post hoc analysis (fig. 1H). Figure 2A was the same as figure 1A. Figure 2, B through E, was the same as figure 1H. Figure 3A had no data analysis. Figure 3, B and C, was the same as figure 1H. Figure 3, D through G, was the same as figure 1D. Figure 4A included no data analysis. Figure 4B was the same as figure 1H. Figure 5, A through C, was the same as figure 1H. Figure 6A was the same as figure 1C. Figure 6, B through F, was the same as figure 1D. Figure 6, G through J, was the same as figure 1H. Significance was set as P < 0.05, and the n value represented experiments with individual rats.

CFA increased the expression of Brd4 in the DRG samples (fig. 1A and table 1; 2.08 ± 0.22, 2.62 ± 0.39, 2.17 ± 0.51, 2.02 ± 0.26, and 1.82 ± 0.25 fold; n = 6) with time-associated reductions in the PWL (fig. 1B; 6.5 ± 3.6, 4.0 ± 1.5, 5.0 ± 1.4, 8.1 ± 3.4, and 9.7 ± 3.4 s; n = 7). At day 1 after the injection (maximal responses), CFA enhanced the immunofluorescence of Brd4 (3.78 ± 0.38 fold; n = 7), which coincided with that of NeuN, NF-200, CGRP, and IB4 (fig. 1C), suggesting that this effect was not restricted to a specific neuronal subpopulation. After confirming that our siRNA procedure efficiently and specifically knocked down Brd4 expression in the DRG (see Supplemental Digital Content 1, https://links.lww.com/ALN/B525) and did not affect the motor response in the naive rats (see Supplemental Digital Content 2, https://links.lww.com/ALN/B525) or the PWL in the saline-treated rats (see Supplemental Digital Content 3, https://links.lww.com/ALN/B525), we observed that the Brd4-specific siRNA (10 μl; 5 μg) reversed the CFA-induced hyperalgesia (fig. 1D; from 4.2 ± 1.4 to 7.5 ± 1.5 s; n = 7) and Brd4 expression in the DRG (fig. 1E; 0.68 ± 0.16 fold; n = 6). Although JQ1, which is a Brd4 inhibition, had no effect on the PWL in the saline-treated rats (see Supplemental Digital Content 4, https://links.lww.com/ALN/B525), JQ1 (10 μl; 10, 30, and 100 μM) dose-dependently attenuated the CFA-induced hyperalgesia on day 1 (fig. 1F; from 4.7 ± 1.8 to 5.0 ± 1.2, 8.2 ± 1.9, and 9.2 ± 1.3 s; n = 7) and ameliorated the hyperalgesia developed on days 5 and 10 (Supplemental Digital Content 5, https://links.lww.com/ALN/B525); however, JQ1 (10 μl, 100 μM) did not affect the expression of Brd4 in the DRG (fig. 1G; 1.02 ± 0.18 fold; n = 6). Furthermore, CFA decreased the total distance traveled, number of rears, and duration of rearing activity (fig. 1H; from 6,004 ± 552 to 2,925 ± 629 cm, 58 ± 12 to 20 ± 7, and 86 ± 30 to 18 ± 10 s; n = 8), which was reversed by the Brd4-specific siRNA (10 μl, 5 μg; 4,655 ± 1,388 cm, 48 ± 21, 72 ± 34 s; n = 8) and JQ1 (10 μl, 100 μM; 4,407 ± 887 cm, 48 ± 28, 67 ± 41 s; n = 8).

In addition to increasing the number of Nav1.7 promoter fragments that were immunoprecipitated by the H3-specific antibodies (fig. 2A; 4.43 ± 0.86 fold; n = 6), CFA upregulated the mRNA and protein expression of Nav1.7 in the DRG samples (1.83 ± 0.15 and 2.75 ± 0.52 fold; n = 6), and this effect was reversed by the Brd4-specific siRNA and JQ1 (fig. 2B, 10 μl, 5 μg, 1.25 ± 0.31 and 1.42 ± 0.19 fold, n = 6; fig. 2C, 10 μl, 100 μM, 1.26 ± 0.47 and 1.60 ± 0.32 fold, n = 6). CFA increased the amount of Brd4 antibody-precipitated Nav1.7 promoter (2.06 ± 0.23 fold; n = 6), which was reversed by the Brd4-specific siRNA and JQ1 (fig. 2D; 10 μl, 5 μg and 10 μl, 100 μM; 0.47 ± 0.48 and 0.83 ± 0.30 fold; n = 6). Finally, the Brd4-specific siRNA (10 μl, 5 μg) markedly reduced the count of the CFA-enhanced Brd4-positive, Nav1.7-positive, and Brd4/Nav1.7 double-labeled neurons (fig. 2E; from 2.90 ± 0.35 to 1.22 ± 0.45 fold, from 4.00 ± 0.51 to 1.39 ± 0.38 fold, and from 3.05 ± 1.11 to 1.10 ± 0.36 fold; n = 7). JQ1 (10 μl, 100 μM) exhibited a similar effect but failed to affect the number of Brd4-positive neurons (3.00 ± 0.67, 1.34 ± 0.44, and 1.26 ± 0.44 fold; n = 7).

On day 1 after the CFA injection, most of the Brd4, CDK9, and Nav1.7 immunoreactivity was colocalized in the DRG slices (see Supplemental Digital Content 6, https://links.lww.com/ALN/B525). The Brd4-, CDK9-, and Nav1.7-specific antibody-labeled immunoreactivity in the Brd4 antibody-recognized precipitation was purified from the DRG of the CFA-treated rats (fig. 3A). CFA increased the abundance of CDK9 antibody-precipitated Nav1.7 promoter (5.12 ± 2.46 fold; n = 6), and this effect was attenuated by the Brd4-specific siRNA and JQ1 (fig. 3B; 10 μl, 5 μg and 10 μl, 100 μM; 2.20 ± 0.36 and 1.84 ± 0.42 fold; n = 6). The Brd4-specific siRNA (10 μl, 5 μg) and JQ1 (10 μl, 100 μM) both decreased the CFA-enhanced expression of CDK9 (fig. 3C; from 1.87 ± 0.28 to 1.10 ± 0.36 and 1.04 ± 0.36 fold; n = 6). After confirming that our knockdown protocol has good efficacy and specificity (see Supplemental Digital Content 7, https://links.lww.com/ALN/B525) with no motor deficits (see Supplemental Digital Content 8, https://links.lww.com/ALN/B525) or altered PWL in the saline-treated rats (see Supplemental Digital Content 9, https://links.lww.com/ALN/B525), we observed that the CDK9-specific siRNA (10 μl; 5 μg) ameliorated the hyperalgesia (fig. 3D; from 3.1 ± 0.7 to 11.1 ± 2.5 s; n = 7) and reduced the Nav1.7 promoter fragments immunoprecipitated with the CDK9-specific (0.24 ± 0.05 fold; n = 6) but not Brd4-specific (0.94 ± 0.19 fold, n = 6) antibody in the CFA-treated rats (fig. 3E). Moreover, without affecting the abundance of the Brd4 protein (0.96 ± 0.17 fold; n = 6), the CDK9-specific siRNA (10 μl; 5 μg) reduced the expression of the CDK9 protein (0.31 ± 0.07 fold; n = 6) and Nav1.7 mRNA/protein (0.60 ± 0.19 and 0.56 ± 0.11 fold; n = 6) in the DRG (fig. 3, F and G).

CFA increased the total Nav currents in the DRG neurons (55 ± 11; n = 4 to 6) that were attenuated by JQ1 and the Brd4- and CDK9-specific siRNA (fig. 4, A and B; 10 μl; 100 μM, 5 μg and 5 μg; 19 ± 9, 21 ± 4, and 20 ± 11 pA/pF; n = 4 to 6). A bath application of ProTx-II largely abolished the recorded current, suggesting that the Nav1.7-dependent currents were enhanced after the CFA injection.

CFA enhanced the Brd4, pRNAPII, Nav1.7, and Brd4–pRNAPII–Nav1.7 triple-labeled immunoreactivity in the DRG (fig. 5A and table 2; 3.09 ± 0.60, 3.71 ± 0.62, 5.78 ± 1.07, and 2.59 ± 0.77 fold; n = 7). Although JQ1 (10 μl, 100 μM) did not affect the count of Brd4-positive neurons (3.13 ± 0.61 fold; n = 7), that of CFA-enhanced pRNAPII, Nav1.7, and Brd4–pRNAPII–Nav1.7 triple-labeled neurons (1.11 ± 0.39, 1.35 ± 0.42, and 1.03 ± 0.31 fold; n = 7) was reduced. CFA enhanced the expression of pRNAPII in the DRG (1.95 ± 0.34 fold; n = 6), and this effect was inhibited by the Brd4- and CDK9-specific siRNA and JQ1 (fig. 5B and table 3; 10 μl; 5 μg, 5 μg, and 100 μM; 1.34 ± 0.25, 1.16 ± 0.22, and 1.32 ± 0.19 fold; n = 6). CFA also increased the abundance of pRNAPII antibody-precipitated Nav1.7 promoter fragments (1.84 ± 0.31 fold; n = 6), which was inhibited by JQ1 and the Brd4- and CDK9-specific siRNA (fig. 5C and table 4; 10 μl; 100 μM, 5 μg and 5 μg; 0.84 ± 0.14, 0.87 ± 0.47, and 0.65 ± 0.43 fold; n = 6).

CFA enhanced the mRNA expression of TNF-α in the DRG (fig. 6A; 1.61 ± 0.38 fold; n = 6). The TNF-α–neutralizing antibody (10 μl, 100 ng) decreased the Nav1.7 promoter that was immunoprecipitated by the Brd4-, CDK9-, pRNAPII-, and H3-specific antibodies (fig. 6, B and C; 0.61 ± 0.11, 0.48 ± 0.08, 0.67 ± 0.08, and 0.19 ± 0.16 fold; n = 6), the Nav1.7 mRNA (fig. 6D; 0.54 ± 0.18 fold; n = 6) and protein (see Supplemental Digital Content 10, https://links.lww.com/ALN/B525) expression in the DRG, and allodynia (fig. 6E; 9.4 ± 0.8 s; n = 7). TNF-α (10 μl, 1 pM) increased the abundance of the Nav1.7 promoter that was immunoprecipitated by the H3-, Brd4-, CDK9-, and pRNAPII-specific antibodies (fig. 6, F and G; 1.44 ± 0.14, 3.61 ± 2.30, 1.83 ± 0.53, 2.08 ± 0.52 fold; n = 7), the abundance of Nav1.7 mRNA (fig. 6H; 1.90 ± 0.29 fold; n = 6) per protein (see Supplemental Digital Content 11, https://links.lww.com/ALN/B525) in the DRG samples, and hyperalgesia (fig. 6I; 3.7 ± 0.6 s; n = 7) in the naive rats; these effects were reduced by the Brd4- (10 μl, 5 μg; 1.20 ± 0.75, 0.96 ± 0.52, 0.84 ± 0.52, 1.01 ± 0.63 fold; 7.7 ± 1.5 s; n = 6 to 7) and CDK9-specific siRNA (10 μl, 5 μg; 1.47 ± 1.13, 0.76 ± 0.19, 0.97 ± 0.37, 0.93 ± 0.68 fold; 10.9 ± 2.5 s; n = 6 to 7). Moreover, TNF-α (10 μl, 1 pM) increased the total Nav1.7 currents (61 ± 8 pA/pF; n = 6) in the DRG neurons that were attenuated by the Brd4- and CDK9-specific siRNA (fig. 6J; both 10 μl, 5 μg; 21 ± 5 and 20 ± 14 pA/pF; n = 6).

In this study, we proposed a Brd4-associated epigenetic mechanism as a previously unknown pathogenic pathway for inflammatory hyperalgesia. Intraplantar CFA injections increased the expression of Brd4 in DRG neurons, which promoted the progression of inflammatory hyperalgesia by binding to the acetylated Nav1.7 promoter, thereby upregulating Nav1.7 transcription and expression. Mechanistically, the expressed Brd4 in the DRG neurons recruits and interacts with CDK9 to phosphorylate RNAPII, which subsequently facilitates the transcription and expression of Nav1.7. Furthermore, we explored the contribution of Brd4 to inflammatory hyperalgesia by focusing on the plasticity that occurs in the DRG, and the potential role of spinal Brd4 in the machinery underlying inflammatory hyperalgesia requires additional elucidation because plastic changes in the dorsal horn have been associated with forms of nociceptive sensitization.^25,31 

In addition, by injecting TNF-α into naive animals and conversely injecting specific neutralizing antibodies into animals experiencing CFA-induced hyperalgesia, we observed that TNF-α–associated inflammatory hyperalgesia involves Brd4-dependent epigenetic processing at the Nav1.7 promoter through the Brd4–CDK9–pRNAPII cascade to upregulate the expression of Nav1.7 in DRG neurons but not exclusively. Notably, in diabetic rats, TNF-α contributes to neuropathic allodynia and hyperalgesia by impacting the nuclear factor (NF)-κB–dependent expression of Nav1.7,³²  and Brd4 binds to acetylated RelA of NF-κB, which recruits CDK9 and phosphorylates RNAPII, which facilitates the transcription of inflammatory genes.¹²  Our published data show that CFA increased the interaction between Brd4 and RelA in the DRG samples (see Supplemental Digital Content 12, https://links.lww.com/ALN/B525); however, the possibility that Brd4 binds to RelA and acts cooperatively with NF-κB to trigger the CDK9–pRNAPII–Nav1.7 cascade to participate in the development of inflammatory hyperalgesia cannot be excluded. Interestingly, the TNF-α–neutralizing antibody markedly decreased the amount of H3Ac–Nav1.7 promoter precipitates compared with antibodies against Brd4–, CDK9–, and pRNAPI–Nav1.7 in promoter precipitates. The possibility that cytokines other than TNF-α contribute to inflammatory hyperalgesia via the Brd4–CDK9–pRNAPII–Nav1.7 cascade requires additional investigation, because cytokine- and chemokine-triggered cellular–systemic responses are associated with pain pathology, and interleukin-1β mediates airway inflammation through Brd4-dependent machinery.³³  However, in addition to H3, Brd4 exerts its effect through interactions with other acetylated histones, such as H4.³⁴  Moreover, as described above, Brd4 binds to acetylated proteins, such as RelA,¹²  to mediate inflammatory responses. Hence, the H3-associated machinery is not only one of the downstream cascades of Brd4-dependent epigenetics. However, the possibility that CFA induced inflammatory hyperalgesia through an H4-dependent pathway cannot be excluded.

By binding to acetylated histone tails, the BET families, including Brd2, Brd3, Brd4, and BrdT,³⁵  participate in transcriptional activation and elongation.^4,5  Cocrystallization analyses revealed that Brd4 has a shape that is complementary to the acetyl–lysine binding cavity,³⁵  which is an important epigenetic marker that regulates gene transcription.³⁶  The Brd4-binding site is closely associated with the expression of target genes,³⁷  and, hence, Brd4 is recognized as a superenhancer of gene expression.³⁸  In addition to demonstrating Brd4 expression in neurons, a recent study has associated neuronal Brd4-dependent gene transcription with the plasticity underlying memory formation.⁶  In addition to regulating learning- and memory-associated neural plasticity in certain brain areas, epigenetic mechanisms specifically regulate hyperalgesia-related plasticity in the DRG, because certain forms of neural plasticity rely on similar molecular machineries.¹  These data prompted us to investigate whether Brd4 plays a role as an expression enhancer in the development of inflammatory hyperalgesia by acting as a reader of acetylated histone markers in DRG neurons. In the present study, we identified that CFA induced Brd4 expression, Brd4 binding to the acetylated Nav1.7 promoter, and Nav1.7 mRNA and protein expression in DRG neurons concomitant with behavioral hyperalgesia. The CFA-induced Nav1.7 protein and mRNA expression, Brd4–Nav1.7 promoter coupling, and hyperalgesia were inhibited by the focal knockdown of Brd4 expression in the DRG and the administration of JQ1, which blocks the BET proteins from binding to acetylated histones.³⁵  In addition to linking the Brd4-associated epigenetic mechanisms to the plasticity underlying inflammatory hyperalgesia, these findings additionally demonstrate that JQ1 suppresses the recruitment of Brd4 by inflammatory genes,⁸  which is consistent with results demonstrated in Helicobacter pylori–induced⁸  and Alzheimer disease–associated inflammation.³⁹  Therefore, our findings revealed a potential therapeutic effect of JQ1 on inflammatory hyperalgesia by limiting Brd4 binding to acetylated histones.

Brd4 transiently binds to acetylated histones, which are rich in transcriptionally active chromatin regions.⁴⁰  Through interacting with acetylated histones H4 and H3, Brd4 activates gene transcription.⁴¹  Importantly, H3 is a residue that is critical for nucleosome stability; thus, Brd4 acetylates H3, resulting in nucleosome eviction and chromatin decompaction.⁴²  Because growing evidence has linked acetyl–histone H3 to the development of chronic hyperalgesia,⁴³  we investigated the involvement of Brd4-dependent epigenetic modifications in nociceptive sensitization by examining H3 acetylation. Consistent with these studies, we observed that the intraplantar CFA injections facilitated histone H3 acetylation at the Nav1.7 promoters in DRG neurons, which was evidenced by the increased abundance of H3-specific, antibody-precipitated Nav1.7 promoter fragments. Nevertheless, the role of other histone families, particularly H4, should be further investigated because histone cross-talk generates a nucleosomal recognition code composed of H3/H4 histones that determine the nucleosome platform for bromodomain protein Brd4 binding.⁴⁴  In addition, H3 phosphorylation promotes H4 acetylation to generate histone cross-talk.⁴⁴ 

Brd4 functions as a critical mediator of transcriptional elongation by recruiting positive transcription elongation factor b (P-TEFb),¹¹  which is a core component of CDK9.⁴⁵  Nevertheless, Brd2, which is another BET family protein with a similar structural composition, did not interact with P-TEFb.¹⁰  Itzen et al.⁴⁶  have shown that Brd4 regulates and stimulates the kinase activity of P-TEFb, which phosphorylates the C-terminal domain (CTD) of RNAPII. Consistently, the CFA injections increased Brd4, CDK9, and pRNAPII expression; Brd4–CDK9–pRNAPII coupling; Brd4–CDK9–pRNAPII–Nav1.7 promoter binding; and Nav1.7 protein–mRNA expression in DRG neurons, which were accompanied by the development of hyperalgesia; these effects were all inhibited by the focal knockdown of Brd4. The pharmacologic antagonism of Brd4 activity by JQ1 markedly decreased CDK9 and pRNAPII protein expression, Brd4–CDK9–pRNAPII binding to the Nav1.7 promoter, and Nav1.7 protein–mRNA expression in the CFA-treated rats. Although the knockdown failed to affect Brd4 expression and Brd4–Nav1.7 promoter binding, focal knockdown of CDK9 expression ameliorated the CFA-induced behavioral hyperalgesia and inhibited the associated CDK9 changes, Brd4–CDK9 coupling, CDK9–Nav1.7 promoter binding, and Nav1.7 protein–mRNA expression in DRG neurons. These findings reveal that the Brd4-dependent epigenetic activation of Nav1.7 gene expression critically contributes to nociceptive sensitization after inflammatory insults through its association with and regulation of CDK9 in DRG neurons.

In addition to acting as an anticancer reagent,⁴⁷  flavopiridol, which is a flavonoid designed to inhibit the activity of CDK1, CDK2, CDK4, CDK6, CDK7, and CDK9,⁴⁸  exhibits therapeutic effects on CFA-induced inflammation.⁴⁹  Moreover, roscovitine, which is another CDK inhibitor that specifically targets CDK1, CDK2, CDK5, CDK7, and CDK9, attenuated CFA-induced thermal hyperalgesia.⁵⁰  Flavopiridol and roscovitine have been shown to be effective against CDK9,⁵¹  and the possibility that these reagents exhibit analgesic effects by impacting other CDK families cannot be excluded. In the present study, we focally knocked down CDK9 expression to elucidate the involvement of CDK9 in inflammation-associated nociceptive sensitization instead of using wide-range CDK inhibitors, such as flavopiridol or roscovitine. Our methodology specifically characterized the role of CDK9 in the pathology of hyperalgesia and avoided complicated pharmacologic effects on members of the CDK families. CDK9 phosphorylates serine residues on the CTD of RNAPII.⁵²  After recruitment to Ser2 at the CTD, CDK9 promotes the transcriptional elongation of targeted genes.⁵³  Moreover, the ectopic expression of Brd4 in HeLa cells resulted in a CDK9-dependent phosphorylation of Ser2 at the CTD.¹⁰  Consistently, in this study, we investigated the epigenetic mechanisms underlying nociceptive sensitization using a selective antibody against pSer2 of RNAPII, and our results showed that CFA-elevated Brd4 expression induced CDK9-dependent Ser2 phosphorylation in the CTD of RNAPII. However, the phosphorylation of RNAPII occurs primarily at the Ser2 and Ser5 residues of the CTD,⁵⁴  and the potential roles of Ser5 phosphorylation⁵⁴  cannot be excluded because Ser2 and Ser5 work cooperatively in gene activation⁵⁵  and Ser5 phosphorylation is confined to promoter regions and is necessary for transcription initiation. In addition, the possible role played by Ser7 phosphorylation in postinflammatory hyperalgesia cannot be excluded, because a previous study has demonstrated that Brd4-activated CDK9 preferentially phosphorylates Ser7 on the CTD of RNAPII.⁴⁶ 

In contrast to systemically blocking histone acetylation of genes that could widely impact protein transcription and hence result in serious side effects, the findings in this study offer a basis for the development of a medical strategy for relieving inflammatory hyperalgesia by spinal application of JQ1, which more specifically and restrictively limits Brd4–histone coupling in the DRG. Nevertheless, the potential side effects of this treatment need additional investigation.

Supported by the Ministry of Science and Technology (Taipei, Taiwan) grants MOST 105-2628-B-715-003-MY3, 104-2320-B-715-004-MY3, NSC 102-2628-B-715-001, 101-2320-B-715-001-MY3, and MOST 105-2320-B-715-003-MY2 (to Drs. Peng and Ho); by Mackay Memorial Hospital (Taipei, Taiwan) grants MMH-MM-10206, MMH-MM-10302, MMH-MM-10403, MMH-MM-10503, and MMH-MM-10608 (to Dr. Peng); and by Department of Medicine, Mackay Medical College (New Taipei, Taiwan) grants 1001A03, 1001B07, 1011B02, 1021B08, 1031A01, 1031B07, 104B06, 1042A08, 1051B03, and 1051B04 (to Drs. Peng, Ho, and Cheng).

The authors declare no competing interests.